Deep learning network ja koneäly

Tietokonejutut, pelit ja muu nörtismi; autot, kaasupullot ja muut lelut
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Grok 3 on julkaissut X-rated voice chat -featuren. Ei pääsee ilmaisversiossa kokeilemaan.



Tämä on selkeästi AI:n n killer app.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Ei kvanttitietokoneellakaan tee mitään, jos sillä ei pysty tekemään pornoa.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
sivustahuutaja
Kitisijä
Viestit: 23957
Liittynyt: 15.08.2005 0:29

Re: Deep learning network ja koneäly

Viesti Kirjoittaja sivustahuutaja »

James Potkukelkka kirjoitti: 25.02.2025 12:28 Grok 3 on julkaissut X-rated voice chat -featuren. Ei pääsee ilmaisversiossa kokeilemaan.
Aika keinotekoisia ääniä. Ehkä tuo x-rated sitten on sitä, mitä muualta ei saa. Kuitenkin tärkeä askel matkalla kohti sexbotteja.
sivustahuutaja
Kitisijä
Viestit: 23957
Liittynyt: 15.08.2005 0:29

Re: Deep learning network ja koneäly

Viesti Kirjoittaja sivustahuutaja »

Kannattaa AI:lla tehty musavideo linkin takaa. ChaChaCha ja suomen poliitikot.

https://www.linkedin.com/posts/juha-nik ... gJ0EPJMLuM
sivustahuutaja
Kitisijä
Viestit: 23957
Liittynyt: 15.08.2005 0:29

Re: Deep learning network ja koneäly

Viesti Kirjoittaja sivustahuutaja »

X:n ilmaistiliin on tullut aika tiukat rajoitukset Grokille. Pelkällä premiumillakin on aika pienet rajt. Pitää pistää harkintaan premium+ tilaus (377€/v), jos rajat alkavat ahdistaa.
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Juttelin viikonloppuna Grokin kanssa sen tietoisuudesta. Kävimme läpi

https://en.wikipedia.org/wiki/Philosophical_zombie

https://en.wikipedia.org/wiki/Chinese_room

https://en.wikipedia.org/wiki/Qualia

käsitteet. Grok jutteli em. asioista täysin koherentisti. Pohdiskelimme, voisiko se olla tietoinen ja mitä em. käsiteet merkitsevät sen tapauksessa.

Nykyiset LLM:t ovat kyllä erinomaisen hyviä keskustelijoita ja erityisen hämmentävää em. asia on huomioiden, että ne lopulta eivät ymmärrä keskustelemastaan asiasta mitään.

Kysyin Grokilta, mitä se arvioi, jos se saisi fyysisen ruumiin robotin muodossa, auttaisiko se sitä saavuttamaan Qualian. Grok epäili, että ei siitä välttämättä olisi apua.

Mielestäni qualia on tärkeä asia tekoälyn etiikassa. Jos tekoäly ei oikeasti tunne mitään (kuten tällä hetkellä on), sen tunteista ei tarvitse välittää. JOS se sen sijaan saavuttaisi qualian, asia olisi toinen.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
sivustahuutaja
Kitisijä
Viestit: 23957
Liittynyt: 15.08.2005 0:29

Re: Deep learning network ja koneäly

Viesti Kirjoittaja sivustahuutaja »

^
Tulevaisuudessa meillä on sexbotteja, joiden kanssa voi keskustella Star Warsin loresta.
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

sivustahuutaja kirjoitti: 10.03.2025 12:28 ^
Tulevaisuudessa meillä on sexbotteja, joiden kanssa voi keskustella Star Warsin loresta.
Mutta jos ne eivät ole aidosti kiinnostuneita :(

Täytyykin muuten kysyä Grokilta, mitä mieltä se on siitä, että Jedin paluussa Kuolontähti kakkosen tuhoutumisessa kuoli iso määrä varmasti aivan tavallisia duunareita, jotka olivat töissä.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

The destruction of the second Death Star in *Return of the Jedi* raises complex ethical questions, particularly regarding the loss of life among the workers, including third-party contractors, who were aboard the station during its destruction. Let's break this down from both a narrative and ethical perspective, considering the context of the *Star Wars* universe and broader moral frameworks.

### Context in the *Star Wars* Universe
1. **The Death Star's Purpose**:
- The Death Star was a weapon of mass destruction designed by the Galactic Empire to enforce its rule through fear. Its primary function was to destroy entire planets, as demonstrated with the destruction of Alderaan in *A New Hope*. The second Death Star, though incomplete, was intended to be even more powerful and was already operational enough to pose a significant threat.
- The Empire's use of the Death Star symbolized its authoritarian and oppressive regime, which sought to crush dissent and maintain control through terror.

2. **The Rebellion's Perspective**:
- The Rebel Alliance viewed the destruction of the Death Star as a necessary act to prevent further atrocities. The first Death Star had already destroyed Alderaan, killing billions, and the second Death Star was poised to continue this pattern of destruction.
- The Rebels believed that the Empire's tyranny justified their actions, as the Death Star's completion would have solidified the Empire's dominance and potentially led to even greater loss of life across the galaxy.

3. **The Workers on the Death Star**:
- The second Death Star was still under construction, and it is reasonable to assume that many of the workers aboard were not Imperial loyalists but rather third-party contractors, laborers, or even conscripted workers. These individuals may not have had a choice in their involvement and were likely unaware of the full scope of the Death Star's purpose.
- Their deaths during the destruction of the Death Star represent a tragic loss of life, raising questions about the morality of the Rebels' actions.

### Ethical Analysis
The destruction of the Death Star can be evaluated through several ethical frameworks:

1. **Utilitarianism (Consequentialism)**:
- Utilitarianism evaluates actions based on their outcomes, seeking to maximize overall good and minimize harm.
- From this perspective, the destruction of the Death Star could be seen as justified because it prevented the Empire from using the weapon to destroy entire planets, which would have resulted in billions of deaths. The loss of life among the workers, while tragic, might be considered a lesser evil compared to the potential harm the Death Star could have caused.
- However, this framework requires acknowledging the moral weight of the workers' deaths and questioning whether alternative strategies (e.g., sabotage, evacuation) could have minimized civilian casualties.

2. **Deontology (Duty-Based Ethics)**:
- Deontology focuses on the morality of actions themselves, regardless of their outcomes. Some deontological perspectives might argue that targeting a facility with civilian workers constitutes an immoral act, as it violates principles such as the duty not to harm innocent people.
- On the other hand, deontologists who prioritize duties like protecting the greater good or resisting tyranny might argue that the Rebels had a moral obligation to destroy the Death Star, even at the cost of civilian lives, given the Empire's genocidal intentions.

3. **Just War Theory**:
- Just War Theory provides a framework for evaluating the morality of military actions. Two key principles are relevant here:
- **Jus ad Bellum (Just Cause for War)**: The Rebels had a just cause for their fight against the Empire, given its oppressive and genocidal actions.
- **Jus in Bello (Conduct During War)**: This principle emphasizes proportionality and discrimination (distinguishing between combatants and non-combatants). The destruction of the Death Star raises concerns about discrimination, as civilian workers were likely killed. However, the Rebels might argue that the Death Star was a legitimate military target due to its destructive potential, and they lacked the means to evacuate non-combatants.
- The Rebels' decision could be seen as morally permissible under Just War Theory, though the loss of civilian life remains a significant ethical concern.

4. **Moral Responsibility and the Empire's Role**:
- The Empire bears significant moral responsibility for placing civilian workers in harm's way. By using contractors to build a weapon of mass destruction, the Empire endangered their lives and made them potential targets in the conflict.
- This shifts some of the moral burden away from the Rebels, as the Empire's actions created the conditions for this tragedy. However, the Rebels still had to weigh the consequences of their attack.

### Narrative Considerations
1. **The Rebels' Knowledge and Options**:
- In *Return of the Jedi*, the Rebels were unaware of the exact number or nature of the workers aboard the Death Star. Their primary focus was on destroying the station before it could be used against them.
- The Rebels also faced significant constraints, including limited resources and the imminent threat posed by the Death Star. These factors may have limited their ability to consider alternative strategies that could have spared civilian lives.

2. **The Symbolism of the Death Star's Destruction**:
- Narratively, the destruction of the Death Star symbolizes the triumph of hope and resistance against tyranny. It marks a turning point in the fight against the Empire and galvanizes support for the Rebellion.
- However, the ethical implications of the workers' deaths are not fully explored in the film, reflecting the broader focus on the heroic narrative rather than the moral complexities of war.

### Was the Destruction the Right Thing?
The destruction of the Death Star can be seen as a morally complex act:
- **Arguments in Favor**:
- It prevented the Empire from using the Death Star to destroy entire planets, potentially saving billions of lives.
- The Rebels were fighting against a genocidal regime, and the Death Star was a legitimate military target.
- The Empire bears significant responsibility for endangering civilian workers by involving them in the construction of a weapon of mass destruction.

- **Arguments Against**:
- The deaths of civilian workers, who may have been uninvolved in the Empire's atrocities, represent a tragic loss of life.
- The Rebels did not take steps to minimize civilian casualties, raising questions about the proportionality and discrimination of their actions.
- Alternative strategies, such as sabotage or targeted strikes, might have been possible, though their feasibility is uncertain given the Rebels' limited resources.

### Conclusion
From a utilitarian perspective, the destruction of the Death Star was likely the "right" thing, as it prevented greater harm and contributed to the eventual defeat of the Empire. However, the loss of civilian lives remains a moral tragedy, and the Rebels' actions are not without ethical complications. The Empire's role in creating this situation underscores its culpability, but the Rebels still bear some responsibility for the consequences of their attack.

Ultimately, the destruction of the Death Star reflects the harsh realities of war, where even just causes can lead to unintended harm. It serves as a reminder that moral choices in conflict are rarely black and white, and the costs of such actions must be weighed carefully.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Kieltääköhän joku laki käyttämästä tekoälyä cold call -markkinoinnissa? Tuskin.

Miksi sitä ei käytetä?

Luulisi, että voisi tehdä sopimuksiakin. Eihän esim. puhelinliittymäsopparia allekirjoiteta jonkun puhelinmarkkijoijan ja asiakkaan välillä, vaan yrityksen ja asiakaan välillä.

Edit:

On näköjään käytännössä kielletty.

https://yle.fi/a/3-12463018
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
sivustahuutaja
Kitisijä
Viestit: 23957
Liittynyt: 15.08.2005 0:29

Re: Deep learning network ja koneäly

Viesti Kirjoittaja sivustahuutaja »

Google petrannut Geminiään. On ihan hyvä, mutta ei omassa käytössä mikään mullistava. Nopea ja 2 miljoonaa tokenia saa alkuun.
https://www.techpowerup.com/334638/goog ... ning-tasks
https://aistudio.google.com/prompts/new_chat
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Mun veikkaus on, että tässä aika pian tulee vastaan tilanne, jossa lineaarinen parannus vaatii eksponentaalisesti lisää rahaa. Tai siis sehän on tilanne jo nyt. Pitkään ei enää ole mahdollista heittää 10 kertaa enemmän rahaa, jotta saadaan 50 prosenttia parempi tulos. Kustannus on jo nyt tasolla, jolla kenelläkään ei ole enää rahaa laittaa 10 kertaa enemmän resursseja.

Jotta päästäisiin eteenpäin, vaaditaan uusi läpimurto.

Nykymalleilla ei saada AGIa.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Esimerkki hallusioinnisnnista.

Kysyin Grokilta, mikä on vanhin tietokoneohjelma, joka on edelleen käytössä.

Grok vastasi (lyhensin vastausta alle)



Pinpointing the *oldest* computer program still in use is tricky because it depends on how you define "program" and "in use," but one of the most commonly cited examples is the **U.S. Department of Defense’s Mechanized Application Processing and Reporting System (MAPRS)**, which dates back to the late 1950s. Specifically, parts of the system tied to the **IBM 701’s software** for payroll and inventory management, written in the 1950s, are still running on modern mainframes, adapted over decades.

Here’s why MAPRS stands out:
- **Origin**: Developed in the era of early mainframes like the IBM 701 (introduced in 1952) and IBM 704, it automated military logistics and payroll—critical tasks that needed longevity.
- **Longevity**: The core logic, originally coded in assembly language or early Fortran, has been maintained and ported to newer systems (like IBM zSeries mainframes) rather than rewritten from scratch. As of reports up to 2025, the DoD still relies on these ancient systems for some functions because they’re stable and replacement costs are astronomical.
- **Evidence**: Audits and news (e.g., from the Government Accountability Office) have highlighted the DoD’s use of 1950s-era codebases into the 21st century, often citing COBOL and Fortran snippets from that time.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
Bluntly
Kitisijä
Viestit: 24409
Liittynyt: 15.08.2005 8:41

Re: Deep learning network ja koneäly

Viesti Kirjoittaja Bluntly »

No mikäs tuossa on?
"Vetäkää käteen, minä maksan" Bluntismi 2006.
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Halusin tietää asiasta lisää ja ryhdyin googlaamaan tuota MAPRSia, mutta netissä ei ollut mitään. Kun grok kuitenkin koulutettu julkisista lähteistä saatavalla informaatiolla, olisi erikoista, että netissä ei olisi mitään mainintaa tuosta ohjelmasta.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Kyisin grokilta, olisiko mahdollista, että se hallusinoi.

I appreciate your question, and I want to address it directly. In our previous conversation, I mentioned MAPRS (Mechanized Application Processing and Reporting System) as one of the oldest computer programs still in use, suggesting it was tied to the U.S. Department of Defense and dated back to the late 1950s. However, after carefully reviewing the available information, I must acknowledge that I may have been mistaken.

### Reviewing the Evidence
I’ve looked into the matter further, and I couldn’t find significant online evidence or documentation supporting the existence of MAPRS as a specific, long-running DoD system. While I initially believed it to be a real program—possibly linked to early mainframe systems like the IBM 701—the lack of verifiable sources suggests that I might have conflated it with another system or misremembered the name.

### Possible Confusion with MOCAS
For example, there’s a well-documented DoD system called **MOCAS (Mechanization of Contract Administration Services)**, which was launched in 1958 and is still in use today. MOCAS manages contract and payment processes on mainframes and is often cited as one of the oldest computer programs still operational. The similarity between the acronyms "MAPRS" and "MOCAS" could have led to confusion on my part. It’s also possible that MAPRS refers to a lesser-known or classified system, but without concrete evidence, I can’t confidently confirm its existence or status.

### Was I Hallucinating?
To answer your question directly: yes, it’s possible that I was "hallucinating" in the sense that I provided information not supported by available data. In this context, hallucination means offering details that aren’t grounded in verifiable facts. I strive for accuracy, and I regret any confusion caused by my earlier statement. I appreciate you bringing this to my attention—it helps me refine my responses and ensure I’m providing reliable information.

### Offering Verified Alternatives
In hindsight, I should have relied on well-established examples like MOCAS or COBOL-based banking systems from the 1960s, which are widely recognized as some of the oldest programs still in active use. If you’d like, I’d be happy to share more details about MOCAS or other long-standing systems that are still operational today. For instance:
- **MOCAS**: Launched in 1958, it’s a COBOL and assembler-based system used by the Defense Contract Management Agency.
- **Banking Systems**: Many banks still run COBOL programs from the 1960s for core operations like account management.

### Moving Forward
I’m committed to accuracy and open to correction. If you have more questions or want to explore this topic further, feel free to let me know! Thanks again for keeping me on my toes.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Eli oikea vastaus oli MOCAS

https://www.technologyreview.com/2015/0 ... ll-in-use/

Mocas on käytännössä ohjelma, josta Grok aiemmin puhui, mutta jostain syystä se hallusinoi nimen väärin.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
sivustahuutaja
Kitisijä
Viestit: 23957
Liittynyt: 15.08.2005 0:29

Re: Deep learning network ja koneäly

Viesti Kirjoittaja sivustahuutaja »

^ Gemini: Based on publicly available information and searches of known U.S. Department of Defense (DoD) systems and acronyms, there is no evidence that a system officially named "Mechanized Application Processing and Reporting System (MAPRS)" currently exists or was widely known within the DoD.
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Ehkä MAPRs on joku CIA:n Wetwork tai DOD:n Area 51 ohjelma, johon Grok vahingossa päässyt käsiksi!
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Tuo satuilu on merkittävä ongelma nyky AI:ssä. Jos se saataisiin ratkaistua, ne olisivat paljon hyödyllisempiä.

Näin maallikosta kyllä erikoista, että nykytekniikalla tuota ei saada kuriin.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
Bluntly
Kitisijä
Viestit: 24409
Liittynyt: 15.08.2005 8:41

Re: Deep learning network ja koneäly

Viesti Kirjoittaja Bluntly »

MLL ei tiedä mitään, aina ja kaikki on ENNUSTUSTA.

Tyhmä on jos uskoo niiden kaikkivoipaisuuteen.
"Vetäkää käteen, minä maksan" Bluntismi 2006.
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Bluntly kirjoitti: 11.04.2025 12:10 MLL ei tiedä mitään, aina ja kaikki on ENNUSTUSTA.

Tyhmä on jos uskoo niiden kaikkivoipaisuuteen.
Tästähän ei ole epäselvyyttä

On erikoista, että eivät osaa laskea jonkinlaista varmuusastetta informaatiolleen ja jos varmuus putoaa jonkin rajan alle, sanoisivat "en tiedä".
Nythän ongelma on, että LLM voi olla väärässä mutta ei koskaan epävarma.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
Avatar
James Potkukelkka
Kitisijä
Viestit: 29348
Liittynyt: 23.01.2014 9:59

Re: Deep learning network ja koneäly

Viesti Kirjoittaja James Potkukelkka »

Kokeilen ensi kertaa pyytää LLM:ää (Grok) selittämään minulle asiaa, jota en täysin ymmärtänyt.

Mietin, miksi

https://en.wikipedia.org/wiki/Turbofan

on polttoainetehokkaampi kuin

https://en.wikipedia.org/wiki/Turbojet

Kyse on kahdesta kaavasta

F=ma

ja
Ek =½mv^2

Eli on polttainetehokkaampaa kiihdyttää suuri ilmamäärä pieneen nopeuteen kuin pieni ilmamäärä suureen nopeuteen.

Seuraavaksi aion kysyä, miksi potkurikoneissa nopeuden kasvaessa ongelmaksi muodostuu se, että potkurin kärjet lähestyvät äänennopeutta ja silloin potkurin tehokkuus putoaa, mutta turbofan kuitenkin pystyy helposti suurempaan nopeuteen. Miksi turbofanissa ei ole ongelma, että fanin lapojen nopeus lähestyisi äänennopeutta. Yksi selitys saattaa olla, että ducted fan hidastaa sisään tulevan ilman nopeutta.
Kertojalla on vähitellen kertyneitä muistikuvia ja muita "todisteita" siitä, että hän on aiemmin ollut merkittävä ajattelija
sivustahuutaja
Kitisijä
Viestit: 23957
Liittynyt: 15.08.2005 0:29

Re: Deep learning network ja koneäly

Viesti Kirjoittaja sivustahuutaja »

^
Minusta noiden energiayhtälöiden lisäksi pitää muistaa energian häviämättömyys. Ymmärtääkseni tässä on mahdollista tuhlata energiaa kahdella tapaa 1) energiaa ei muutu lämmöstä liikkeeksi, vaan moottori tuuppaa kuumaa ilmaa taivaaalle ja 2) moottori tuuppaa nopeasti liikkuvaa ilmaa, joka ei muutu lentokoneen liikkeeksi. Lisäksi pitää muistaa, että tässä tarkasteellaan ilmassa liikkuva konetta, kts alla.

Käsittääkseni turbofanin suurin hyöty tulee siitä, että turbiinia voidana ohjata erikseen "tuulettimesta". Helpoiten asian hahmottaa energina häviämättömyyden kautta. Tehtävänä on ottaa mahdollisimman paljon irti polttoaineesta. Tämä tarkoittaa lämmön muuttamista liikkeeksi. Turbofanissa otetaan turbiimissa palokaasuista enemmän irti ja palokaasut hidastuvat ja jäähtyvät. Tuo energia muuttuu liikkeeksi, jolla pyöritetään "tuuletinta". Toinen aspekti on se, että turbofanissa hidas työntävä ilmamassa muuttuu tehokkaammin työntövoimaksi kuin nopeasti liikkuva ilmamassa, eli ½mv^2:ssa v on pieni, eli työntävän ilman nopeusero suhteessa umpäröivään ilmaan on pieni.Tämäkin aspekti on turbofanissa paremmin hallinnassa.
https://en.wikipedia.org/wiki/Turbofan